We should add formats support for GermEval-2014 to OpenNLP, then the existing tools can be used to train OpenNLP on their data files.
I opened a jira for it: https://issues.apache.org/jira/browse/OPENNLP-976 Jörn On Mon, Feb 6, 2017 at 1:10 PM, Richard Eckart de Castilho <[email protected]> wrote: > On 06.02.2017, at 12:43, Martin Wunderlich <[email protected]> wrote: > > > > 2) How come there are no NER models for German (yet)? Anyone working on > them? > > If you want to create a German NER model, you could e.g. use the > GermEval-2014 > dataset. There is an example of how to train an OpenNLP NER model using > DKPro Core > from that dataset rendered as a JUnit test in DKPro Core: > > https://github.com/dkpro/dkpro-core/blob/master/dkpro- > core-opennlp-asl/src/test/java/de/tudarmstadt/ukp/dkpro/core/opennlp/ > OpenNlpNamedEntityRecognizerTrainerTest.java > > It is pretty straight-forward. > > Cheers, > > -- Richard
